Re: We really ought to do something about O_DIRECT and data=journalled on ext4 - Mailing list pgsql-hackers

From Tom Lane
Subject Re: We really ought to do something about O_DIRECT and data=journalled on ext4
Date
Msg-id 6815.1291687813@sss.pgh.pa.us
Whole thread Raw
In response to Re: We really ought to do something about O_DIRECT and data=journalled on ext4  (Robert Haas <robertmhaas@gmail.com>)
Responses Re: We really ought to do something about O_DIRECT and data=journalled on ext4  (Josh Berkus <josh@agliodbs.com>)
Re: We really ought to do something about O_DIRECT and data=journalled on ext4  (Josh Berkus <josh@agliodbs.com>)
List pgsql-hackers
Robert Haas <robertmhaas@gmail.com> writes:
> On Mon, Dec 6, 2010 at 9:04 PM, Josh Berkus <josh@agliodbs.com> wrote:
>> Actually, on OSX 10.5.8, o_dsync and fdatasync aren't even available.
>> From my run, it looks like even so regular fsync might be better than
>> open_sync.

> But I think you need to use fsync_writethrough if you actually want durability.

Yeah.  Unless your laptop contains an SSD, those numbers are garbage on
their face.  So that's another problem with test_fsync: it omits
fsync_writethrough.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: profiling connection overhead
Next
From: Josh Berkus
Date:
Subject: Re: We really ought to do something about O_DIRECT and data=journalled on ext4